Big Ship Little Boat, who Gives way

[Captain Bill]

Nigel,
One more question. How close do you think a small 5-7 knot boat should wait before deciding that a large 15-20 knot vessel on a collision course and is the giveway vessel is not going to maneuver around you and start taking evasive action? I'm thinking here of a open ocean situation, not in established traffic lanes.

[Watermann]

Originally Posted by nigel1

I can understand concerns of ships not sighting a sail boat at sea, and I can put forward a number of reason's for this (please note, this is not an excuse for not keeping a proper lookout)

4) Overload of paperwork and other administrative tasks on OOW. Combine this with smaller crew sizes, OOW will, if allowed to, undertake these tasks when they should be keeping a proper watch.

This is my (and many of my shipmates) number one concern. The administrative tasks expected of bridge watchstanders in today's regulatory reality is out of control, especially in oil trades. How can you expect quality watchstanding when there are things to do on the bridge that have absolutely nothing to do with navigation?

It reminds me of a anecdote I heard or read once. The captain comes up to the bridge and sees a young 3/M working on the computer and says "Twenty years ago, I would've fired you for being on that thing. Today, I'd fire you if you weren't."

Such is the world of modern seafaring...

[Dockhead]

Originally Posted by skipmac

It did occur to me as I was posting the previous that I would be complying with the Colregs by avoiding a collision by taking action if the giveway vessel did not give way. So maybe I am following Colregs more strictly than I think.

I have to say, these discussions are convincing me that I might need to add AIS. Maybe not absolutely necessary for most places I will be cruising but for some areas it could be essential. $#!@#$!!! Another boat buck down the drain.

That being said, back in the good old days, I thought I was doing OK avoiding other vessels by taking frequent bearings on them and acting appropriately, not just going by feel or intuition. Maybe the big boys had already seen me miles and miles away and had acted well before I knew they were there, but regardless, by the time I saw them and started tracking, 99% of the time it became quickly obvious that the other vessel was passing ahead, astern or well abeam of me. The balance of the time a few more bearings as we closed would usually show safe passing with rare exceptions.

All this brings me to something that always bothers me about these collision discussions. Again, my opinion, 99.99% of the collisions occur because someone is, metaphorically speaking, asleep at the wheel. Just to see if I am crazy I did the math and would like to hear if others agree with my assessment of the situation.

The max beam I'm aware of on a big ship is around 200'. A sailboat at 6 kts covers about 10'/second. So, ignoring any margin of error, it would take a typical sailboat 20 seconds to cross completely across the bow of a giant tanker. That tanker at +/- 20 kts might be covering 30-40'/second so in that 20 seconds it takes a sailboat to go 200' the ship would travel roughly 800', round up for that margin of error and call it 1000'.

So for a collision to happen, a sailboat would have to try to cross all the way across the bow of a giant ship going at max speed when the sailboat was less than 1000' in front of that ship.

So again, my thoughts, a collision between a sailboat and a ship is almost 100% avoidable if the sailboat captain is paying attention. Note!! This doesn't mean that the sailboat would not be disrupting traffic, violating Colregs, causing heart attacks on the bridges of multiple commercial vessels or possibly contributing to potential collision situations between other vessels trying to dodge the sailboat. Also, if the other vessel is a high speed ferry or similar then that's a totally different situation.

So, ignoring all the above issues and just focusing on the risk of actual collision, am I wrong to think that risk to be very, very small?

Very interesting line of thought! These kinds of questions lead to better understanding for everyone

First of all -- taking frequent bearings on vessels of concerned is exactly the correct procedure if you don't have AIS. Identifying a collision course from a safe distance is Job 1 in collision avoidance, and skillful use of a hand bearing compass will give you a reasonable idea.

Second -- taking action when the give-way vessel fails to do so it also absolutely correct procedure, required by the Colregs. The problem -- the huge problem with this is that it is devilishly difficult to determine whether the give-way vessel has actually taken action, or not. I don't think many sailors can distinguish a 0.5 mile CPA from a collision course, using a hand bearing compass, no matter how skillful they are. Certainly I cannot. So that means that when you decide to take action, there is a good chance that the give way vessel actually is maneuvering. You can't detect his turn, so you may be turning into danger. There is not much you can do about this except be very diligent with the HBC and make your decision early enough that the other vessel has time to turn back once again if you turned the wrong way or made the wrong decision.

In my opinion, most sailors think in much shorter distances than commercial bridges do, and take action too late. We should ask Nigel, but I think that much action is taken practically over the horizon (so we are not even aware of it), and probably the latest decision point for a commercial bridge is 5 miles or so. I am afraid a lot of sailors bob around out there not even aware of ships around until they are closer than the ships' decision points, so a dodge at that late stage can really screw everything up. But the way you describe what you do, sounds all correct.


Now to the question of "getting out of the way". If you had perfect knowledge of where your course is going to intersect with the course of the ship, and you have a perfect solution for a new course which increases your CPA by one foot for every foot you travel, then your calculations would be correct. HOWEVER, reality will be very different. Even with AIS, you don't have perfect information about the other vessels course, and with a HBC it's very imprecise. So your intersecting courses are not perfect mathematical lines, but cones of uncertainty, narrow or wider cones depending on the quality of your data. You are not safe until you get outside that cone. ON TOP OF THAT, a safe passing distance is not one foot from the ship's side. A cable is already terrifyingly close, and still somewhat dangerous.

So you need to make enough distance with your maneuver to overcome all of these things:

1. Uncertainty of where the ship will be when your courses intersect
2. Minimally safe passing distance, say a cable, at least.
3. The inefficiency of dodging on a non-ideal course, which means you are not getting a foot of useful separation out of every foot you travel.

I'm not even mentioning the fact that you don't know for sure whether he is on a steady course or not -- he may well be turning (he ought to be, actually), in which case that cone of uncertainty can be even much wider, dramatically so.

I'm also not mentioning the fact that you will usually not know which way to dodge -- all you can detect is non-changing bearing, but your data is not accurate enough to tell you whether he is in fact passing some distance ahead or behind. You have no way of knowing to dodge this way or that, and if you dodge wrong, you can end up worse off than before you dodged at all. This is a big problem in a head-on approach; less so, obviously at a right-angles crossing.

AND -- even if you do guess correctly which way you need to turn, don't forget you have to execute that turn, before you get on a course which even starts to create separation. That also takes time, time which is evaporating rapidly as the vessel approaches at 20 knots.

All that means that you need a lot more time on your "dodge" to get out of danger, than the mathematically ideal situation you wrote about.


1000 feet is 300 meters -- so 1.5 cables? I think you're likely dead at 1.5 cables on an intersecting course with a vessel travelling 20 knots. You cannot get out of even a very narrow cone of uncertainty about where that vessel will be just a few seconds later.

Before I had AIS, my rule was I would not pass closer than 2 miles ahead of a fast-moving ship, nor closer than 1 mile to either side. I had radar to accurately judge the range -- extremely valuable -- but radar is carp for bearings, so I used a hand bearing compass like you do for that. Based on that rule my in extremis decision point was usually about two miles away. You have to be really diligent with the HBC to get usable data -- many data points to average out the outlying bumps, which takes valuable time and attention.

At two miles or three miles out, I could make a 90 degree-ish turn onto a reciprocal heading to the ship and usually stop the decrease of range -- that is, freeze the CPA -- and let the ship pass by before tacking back onto my course. HOWEVER, even this maneuver is capable of screw-ups, because the crossing angle might not be as perpendicular as it looks, and you might not get the reciprocal heading right by feel. Easier at night when you can clearly see the aspect of the ship from nav lights. With radar, you can check the correctness of your course by watching the range of the ship -- you can see whether you stopped the decrease of range, and if not, you can correct your course.

I know for sure that I was much cursed by ships' bridges for these maneuvers, when they had calculated and executed a course for a smooth crossing, and then I just unexpectedly tacked away, possibly requiring further maneuvering on their part. But without accurate data, I had no choice. AIS is a total revolution.

[s/v Beth]

AIS has become the useful tool in shipping lanes around Or and Wa, I could not imagine going back to the old days of not having one. Ouestion for the big boys: do your instruments pick us up when we paint you with our radar? Does anything show that there is a small boat out there because he has his radar on?

[jackdale]

Joining the discussion rather late.

Nigel - thanks for the post that started the discussion. I have only ever had one issue with a large vessel.

I was broad reaching at night down Juan de Fuca staying north of the TSS. At one point I noticed that we just crossed the boundary so we gybed. Shortly thereafter I heard a ocean going contacting Seattle traffic asking for permission to enter the median as there a "meandering fishing vessel in the TSS" near my position. Since I saw no other vessel visually or on radar I assumed he meant us. I had been monitoring the radar and see the ocean going well off. I contacted him to let him know that I had gybed back out of the TSS and that I was far enough off as not be impeding. He asked if I heard him on 16 or 13. I had heard him hail a fishing vessel and I dual scan 16 and VTS so I did not hear him on 13. I was reminded by Seattle traffic not to impede.

What struck me afterward was that he could not recognize the lights on a sailing vessel. Ihad the sidelights on, no tricolour on the vessel. I am assuming he could see my lights as I could see his sidelights and both masthead lights.

I have contacted vessels when doing passages and have had success 90% of the time. In one instance we had a ocean going crossing from port to starboard and the owner woke me up to get some advise. I used the EBL to verify that he was crossing ahead and the was confirmed by the other vessel. The onwer wanted to alter course by bearing way which would have put us on a collision course with theother vessel. It took me quite some time to convince him that he could take off a lot of boat speed by luffing up. That would have pointed us toward the other vessel, but put us way behind his stern. Eventually I convinced to hold his course and monitor the other vessel. They passed well ahead.

I have also contacted ocean fishing vessels to verify that their nets were deep enough that I could sail between their buoys. They were surprised to see sail boats.

[goboatingnow]

Is it me or where English is widely spoken like in US /Canadian waters there more sail to ship VHF comms. Over this side I don't find it all that useful. ( imagine trying to talk to Spanish trawlers !!!! Lol)

Dave

[Seaworthy Lass]

Originally Posted by Dockhead

At two miles or three miles out, I could make a 90 degree-ish turn onto a reciprocal heading to the ship and usually stop the decrease of range -- that is, freeze the CPA -- and let the ship pass by before tacking back onto my course. HOWEVER, even this maneuver is capable of screw-ups, because the crossing angle might not be as perpendicular as it looks, and you might not get the reciprocal heading right by feel. Easier at night when you can clearly see the aspect of the ship from nav lights.

An alternative to changing direction is simply slowly the boat down. I don't think this is considered as frequently as it should. Moving onto the reciprocal heading of the ship may mean a huge change in course and an unnecessary deviation from your intended path, slowing your journey more than simply slowing the boat would have done.

[boatman61]

Originally Posted by Seaworthy Lass

An alternative to changing direction is simply slowly the boat down. I don't think this is considered as frequently as it should. Moving onto the reciprocal heading of the ship may mean a huge change in course and an unnecessary deviation from your intended path, slowing your journey more than simply slowing the boat would have done.

.........

[jackdale]

Originally Posted by goboatingnow

Is it me or where English is widely spoken like in US /Canadian waters there more sail to ship VHF comms. Over this side I don't find it all that useful. ( imagine trying to talk to Spanish trawlers !!!! Lol)

Dave

Hi Dave

I think in Europe there is a reluctance to use VHF in collision situations; a "vhf assisted collisions" search makes for some interesting discussions.

[jackdale]

Originally Posted by Seaworthy Lass

An alternative to changing direction is simply slowly the boat down. I don't think this is considered as frequently as it should. Moving onto the reciprocal heading of the ship may mean a huge change in course and an unnecessary deviation from your intended path, slowing your journey more than simply slowing the boat would have done.

+2

Alter speed, alter course, or alter speed and course.

[Captain Bill]

I have been thinking of getting a Standard Horizon 2100 with integrated AIS. I like the feature where you can call a ship on showing on the AIS via DSC. The only problem I have is they seem to have forgotten a possibly useful feature. The only option on the radio is to call the AIS target. I think it would also be useful to have the option of doing a position send to the AIS target so that your position showed up on the vessel's ECDIS display. Not quite as good as a full blown AIS transceiver, but possibly useful to a large vessel trying to avoid you. It would also not have the issue of timing and collisions(radio collisions, not ship) that can sometimes be a problem for class B in high traffic areas. In reading the manual you would have to manually enter the MMSI number of the ship in question, then transmit the position send rather than simply picking the ship off of the display. I was thinking of writing SH and making a feature request for their next generation, but would be interested in getting feedback from the forum on how useful such a feature would be.

[boatman61]

Originally Posted by jackdale

+2

Alter speed, alter course, or alter speed and course.

Unless in Fog...

[skipmac]

Originally Posted by Dockhead

Very interesting line of thought! These kinds of questions lead to better understanding for everyone

Thanks. I would certainly hate to add to the confusion.

First of all -- taking frequent bearings on vessels of concerned is exactly the correct procedure if you don't have AIS. Identifying a collision course from a safe distance is Job 1 in collision avoidance, and skillful use of a hand bearing compass will give you a reasonable idea.

Second -- taking action when the give-way vessel fails to do so it also absolutely correct procedure, required by the Colregs. The problem -- the huge problem with this is that it is devilishly difficult to determine whether the give-way vessel has actually taken action, or not.

Usually true but I have often detected course changes by the other vessel by the change of angle on the bow when the ship was close enough.

I don't think many sailors can distinguish a 0.5 mile CPA from a collision course, using a hand bearing compass, no matter how skillful they are. Certainly I cannot.

Nor I, so at times I'm sure I over reacted or reacted too late or too soon for the other vessel. However I would always try make a course change very obvious and hold steady so the other vessel could tell my intentions.

So that means that when you decide to take action, there is a good chance that the give way vessel actually is maneuvering. You can't detect his turn, so you may be turning into danger. There is not much you can do about this except be very diligent with the HBC and make your decision early enough that the other vessel has time to turn back once again if you turned the wrong way or made the wrong decision.

In my opinion, most sailors think in much shorter distances than commercial bridges do,

Which I'm guessing is why, in my experience, when offshore close approaches were rare.

and take action too late. We should ask Nigel, but I think that much action is taken practically over the horizon (so we are not even aware of it), and probably the latest decision point for a commercial bridge is 5 miles or so. I am afraid a lot of sailors bob around out there not even aware of ships around until they are closer than the ships' decision points, so a dodge at that late stage can really screw everything up. But the way you describe what you do, sounds all correct.


Now to the question of "getting out of the way".

I should have qualified the calculations I made as very simplistic example, focusing just on an ideal sailboat crossing the bow of a ship at 90 degrees scenario and ignoring many variables and variations. Just wanted to throw out the most basic calculation and absolutely NOT making any suggestion about how one should handle that situation or how close one should or could safely approach a bit ship. So people don't try that at home!

If you had perfect knowledge of where your course is going to intersect with the course of the ship, and you have a perfect solution for a new course which increases your CPA by one foot for every foot you travel, then your calculations would be correct. HOWEVER, reality will be very different. Even with AIS, you don't have perfect information about the other vessels course, and with a HBC it's very imprecise. So your intersecting courses are not perfect mathematical lines, but cones of uncertainty, narrow or wider cones depending on the quality of your data. You are not safe until you get outside that cone. ON TOP OF THAT, a safe passing distance is not one foot from the ship's side. A cable is already terrifyingly close, and still somewhat dangerous.

So you need to make enough distance with your maneuver to overcome all of these things:

1. Uncertainty of where the ship will be when your courses intersect
2. Minimally safe passing distance, say a cable, at least.
3. The inefficiency of dodging on a non-ideal course, which means you are not getting a foot of useful separation out of every foot you travel.

I'm not even mentioning the fact that you don't know for sure whether he is on a steady course or not -- he may well be turning (he ought to be, actually), in which case that cone of uncertainty can be even much wider, dramatically so.

I'm also not mentioning the fact that you will usually not know which way to dodge -- all you can detect is non-changing bearing, but your data is not accurate enough to tell you whether he is in fact passing some distance ahead or behind. You have no way of knowing to dodge this way or that, and if you dodge wrong, you can end up worse off than before you dodged at all. This is a big problem in a head-on approach; less so, obviously at a right-angles crossing.

AND -- even if you do guess correctly which way you need to turn, don't forget you have to execute that turn, before you get on a course which even starts to create separation. That also takes time, time which is evaporating rapidly as the vessel approaches at 20 knots.

All that means that you need a lot more time on your "dodge" to get out of danger, than the mathematically ideal situation you wrote about.

Yes, absolutely. Mathematical ideals and the real world often/usually are very different.


1000 feet is 300 meters -- so 1.5 cables? I think you're likely dead at 1.5 cables on an intersecting course with a vessel travelling 20 knots. You cannot get out of even a very narrow cone of uncertainty about where that vessel will be just a few seconds later.

Never been anywhere near that close to see and don't want to find out!


Before I had AIS, my rule was I would not pass closer than 2 miles ahead of a fast-moving ship, nor closer than 1 mile to either side. I had radar to accurately judge the range -- extremely valuable -- but radar is carp for bearings, so I used a hand bearing compass like you do for that. Based on that rule my in extremis decision point was usually about two miles away. You have to be really diligent with the HBC to get usable data -- many data points to average out the outlying bumps, which takes valuable time and attention.

At two miles or three miles out, I could make a 90 degree-ish turn onto a reciprocal heading to the ship and usually stop the decrease of range -- that is, freeze the CPA -- and let the ship pass by before tacking back onto my course. HOWEVER, even this maneuver is capable of screw-ups, because the crossing angle might not be as perpendicular as it looks, and you might not get the reciprocal heading right by feel. Easier at night when you can clearly see the aspect of the ship from nav lights. With radar, you can check the correctness of your course by watching the range of the ship -- you can see whether you stopped the decrease of range, and if not, you can correct your course.

I know for sure that I was much cursed by ships' bridges for these maneuvers, when they had calculated and executed a course for a smooth crossing, and then I just unexpectedly tacked away, possibly requiring further maneuvering on their part. But without accurate data, I had no choice. AIS is a total revolution.


I think when I add AIS it will be similar to the learning experience I had when I added radar. Instead of we're 3-4 miles off the coast I found out that we were almost exactly 5 1/2 miles off the coast. Instead of about a mile to the next marker it was 0.75 miles to that marker. I found my guestimates were often pretty far off.

I do have to make one confession. Over the years I've sailed through NY Harbor, up and down the FL coast, through a few choke points like the Straights of Florida, the Old Bahamas Channel and the Windward Passage so I thought I had experienced some heavy ship traffic. Then I looked at the web site showing the AIS traffic in the English Channel and freaked. Based on what I saw I am just a babe in the woods.

I would recommend every sailor going offshore or boating in a high traffic area or around a commercial port go check out Live Ships Map - AIS - Vessel Traffic and Positions. By the way, I am getting no commissions on the increased sale of AIS systems.

[Dockhead]

Originally Posted by Seaworthy Lass

An alternative to changing direction is simply slowly the boat down. I don't think this is considered as frequently as it should. Moving onto the reciprocal heading of the ship may mean a huge change in course and an unnecessary deviation from your intended path, slowing your journey more than simply slowing the boat would have done.

Yes, a good point, and I used to do this sometimes.

However, this maneuver has few disadvantages:

1. It's not as obvious to the other vessel
2. It is not very effective if the angle of your courses is more than 90 degrees; and
3. On the same heading and with way taken off you are unprepared to take further action if you are not achieving enough separation.

None of which matters much if distance is still great enough and angles are right, but the recip course is a more powerful and flexible maneuver for other cases.

If after a tack onto a recip course you are not achieving enough separation, you simply bear away. You still have way on and you're already on a safer course. I don't like bobbing around with way off and big ships about.

[Dockhead]

Originally Posted by Captain Bill

I have been thinking of getting a Standard Horizon 2100 with integrated AIS. I like the feature where you can call a ship on showing on the AIS via DSC. The only problem I have is they seem to have forgotten a possibly useful feature. The only option on the radio is to call the AIS target. I think it would also be useful to have the option of doing a position send to the AIS target so that your position showed up on the vessel's ECDIS display. Not quite as good as a full blown AIS transceiver, but possibly useful to a large vessel trying to avoid you. It would also not have the issue of timing and collisions(radio collisions, not ship) that can sometimes be a problem for class B in high traffic areas. In reading the manual you would have to manually enter the MMSI number of the ship in question, then transmit the position send rather than simply picking the ship off of the display. I was thinking of writing SH and making a feature request for their next generation, but would be interested in getting feedback from the forum on how useful such a feature would be.

You just call the ship by name. Very effective! I'm a big fan of non-emergency DSC functions, but I wouldn't faff around with it in a potential collision situation.

To give the ship your posit, get a transponder! He needs it before you see him, not after there's already a problem! The cost difference is now so trivial that I can think of no reason to have a receive-only AIS.